KR20090021630A - IC estimation method and IC reduction equalizer - Google Patents

Abstract

An ICI reduction equalizer according to a preferred embodiment of the present invention includes a channel estimator, a channel calculator, an ICI estimator, a subtractor, and an equalizer. The channel estimator estimates a channel response from the received signal. The channel calculator calculates a basic component of the channel response and a fluctuating component of the channel response from the estimated channel response. The ICI estimator multiplies a variation component of the channel response by a received signal in a frequency domain, filters the multiplication result according to filter coefficients, and based on the filtering result, an inter-carrier interference (ICI) included in the received signal. Estimate the component. The subtractor subtracts the ICI component from the received signal in the frequency domain. The equalizer equalizes the output signal of the subtractor based on the basic component of the channel response.

Description

Method of estimating Inter-Carrier Interference and ICI mitigating equalizer

The present invention relates to an ICI estimation method and an ICI reduction equalizer, and more particularly, to an ICI estimation method and an ICI reduction equalizer for estimating an ICI (Inter-Carrier Interference) component by approximating a channel to an M1 order model.

Currently, an orthogonal frequency division multiplexing (OFDM) system is applied to various broadcasting / communication systems. An OFDM system is a broadcast / communication system for transmitting data using a plurality of sub-carriers having orthogonality with each other.

1 is a diagram illustrating a transmitter and a receiver in an OFDM system.

In FIG. 1, the OFDM transmitter 110 includes an encoder 111, a mapper 112, an inverse fast fourier transform (IFFT) block 113, a cyclic prefix inserter 114, an RF transmitter 115, and a transmit antenna. 116. In FIG. 1, the OFDM receiver 120 includes a reception antenna 121, an RF receiver 122, a CP remover 123, a fast fourier transform (FFT) block 124, an equalizer 125, and a demapper 126. And a decoder 127.

를 출력한다. In the OFDM transmitter 110, a cyclic prefix (CP) is inserted into an IFFT-transmitted transmission signal to prevent inter-symbol interference (ISI) and to estimate a channel. The CP-inserted transmission signal S _i (n) is output to the wired / wireless channel via the RF transmitter 116 and the transmission antenna 116. The received signal r _i (n) from which the CP is removed at the OFDM receiver 120 is converted into the received signal R _i (k) in the frequency domain by the FFT block 124. The equalizer 125 equalizes the received signal R _i (k) in the frequency domain and estimates the transmitted signal.

Outputs

In order to accurately estimate the transmission signal in the OFDM receiver 120, it is necessary to accurately understand the response characteristics of the channel. However, since the response characteristics of the channel may change with time due to time-selectivity, and the response characteristics of the channel may vary from frequency to frequency due to frequency-selectivity, the response characteristics of the channel may be precisely determined. It's hard to figure out. Meanwhile, in a mobile reception environment, time-selectivity of a channel and frequency-selectivity of a channel impair orthogonality between subcarriers, resulting in ICI (Inter-Carrier Interference). . Although the response characteristics of the channel are accurately determined using pilots included in the cyclic prefix, the estimation of the transmission signal may be inaccurate due to the influence of the ICI.

ICI further complicates the OFDM receiver's accurate estimation of the transmitted signal. In order for the OFDM receiver to accurately estimate the transmission signal, first the ICI component included in the received signal must be estimated, and the ICI component included in the received signal must be removed or reduced based on the estimation result. In order to remove or reduce the ICI component included in the received signal, the amount of computation and computational complexity increases, which complicates the structure of the OFDM receiver. The OFDM receiver is designed in consideration of the trade-off relationship between the accuracy of the estimation and the computational complexity.

An object of the present invention is to provide an ICI estimation method for estimating ICI (Inter-Carrier Interference) components by approximating a channel with an M1 order model. In addition, the present invention is to provide an ICI reduction equalizer for removing or reducing the ICI component estimated by the ICI estimation method.

An ICI reduction equalizer according to a preferred embodiment of the present invention includes a channel estimator, a channel calculator, an ICI estimator, a subtractor, and an equalizer. The channel estimator estimates a channel response from the received signal. The channel calculator calculates a basic component of the channel response and a fluctuating component of the channel response from the estimated channel response. The ICI estimator multiplies a variation component of the channel response by a received signal in a frequency domain, filters the multiplication result according to filter coefficients, and based on the filtering result, an inter-carrier interference (ICI) included in the received signal. Estimate the component. The subtractor subtracts the ICI component from the received signal in the frequency domain. The equalizer equalizes the output signal of the subtractor based on the basic component of the channel response.

In one embodiment of the present invention, the basic component of the channel response is a channel response component that does not vary during one symbol interval, and the variation component of the channel response is a channel response component that varies during one symbol interval. In the case of approximating a channel to an M1-order model, the channel calculator may output the first order variation component of the channel response or the first order variation component of the channel response to the ICI estimator.

 In one embodiment of the present invention, the ICI estimator includes: a multiplier for multiplying a variation component of the channel response by a received signal in the frequency domain and outputting a multiplication result; A filter bank for filtering the multiplication result according to the filter coefficients and outputting the filtering result; And an adder configured to sum the filtering results and output the ICI component.

In one embodiment of the present invention, when approximating a channel to an M1-order model, the multiplier receives the first order variation component of the channel response and the frequency domain output from the channel calculator. A first multiplier for multiplying the signal and outputting a first multiplication result; And a M1 multiplier for outputting a M1 multiplication result by multiplying the M1 difference variation component of the channel response output from the channel calculator and the received signal in the frequency domain.

In one embodiment of the present invention, in the case of approximating a channel to an M1-order model, the filter bank is configured to convert the first multiplication result output from the first multiplier of the multiplier to the first filter coefficients. A first filter for filtering according to and outputting a first filtering result; And a M1 filter outputting the M1 filtering result by filtering the M1 multiplication result output from the M1 multiplier of the multiplication unit according to the M1 filter coefficients. The first to M1 filters may be linear time-invariant filters. The first to M1 filters may be finite impulse response (FIR) filters.

In one embodiment of the invention, the equalizer may be implemented as a one-tap equalizer.

The ICI reduction equalizer according to one embodiment of the present invention can be applied to a receiver of an orthogonal frequency division multiplexing (OFDM) system.

An ICI estimation method according to a preferred embodiment of the present invention comprises the steps of: calculating a first order variation component of a channel response to a first order variation component of a channel response from a channel response estimated from a received signal; Multiplying each of the first order variation component of the channel response to the first order variation component of the channel response by a received signal in a frequency domain to output a first multiplication result to a M1 multiplication result; Filtering each of the first multiplication result to the M1 multiplication result according to each of the first filter coefficients to the M1 filter coefficients and outputting the first to M1 filtering results; And estimating an inter-carrier interference (ICI) component included in the received signal by summing the first filtering result to the M1 filtering result.

In the case of approximating a channel to an M1-order model, the channel response estimated from the received signal is a basic component that does not change during one symbol period and the first variable that varies during one symbol period. It may be classified into a first-order fluctuating component or the M1th-order fluctuating component.

ICI reduction equalizer according to another embodiment of the present invention, the channel estimator for estimating the channel response from the received signal; A channel calculator for calculating a basic component of the channel response and a first-order fluctuating component of the channel response from the estimated channel response; A first multiplier for multiplying a first order variation component of the channel response by a received signal in a frequency domain to output a first multiplication result; A first filter outputting an inter-carrier interference (ICI) component included in the received signal by filtering the first multiplication result according to finite impulse response (FIR) according to first filter coefficients; A subtraction unit for subtracting the ICI component from the received signal in the frequency domain; And an equalizer for equalizing the output signal of the subtractor based on the basic component of the channel response.

The channel calculator can approximate a channel to a first-order linear model. The channel calculator may output a channel response estimate value in the current symbol period as a basic component of the channel response. In addition, the channel calculator subtracts the channel response estimate in the previous symbol interval from the channel response estimate in the next symbol interval, dividing the subtraction result by the channel response estimate in the current symbol interval, The division result can be output as the first order variation component of the channel response.

The channel calculator may include: a first delayer configured to delay and output a channel response estimate value output from the channel estimator for one symbol period; A second delayer for delaying and outputting the output signal of the first delayer for one symbol period again; A subtractor which subtracts the channel response estimate value in the previous symbol interval output from the second delay unit from the channel response estimate value in the next symbol interval output from the channel estimator; And subtracting the subtraction result output from the subtractor by the channel response estimate value in the current symbol interval output from the first delayer, and outputting the division result as the first order variation component of the channel response to the first multiplier. A divider may be provided.

The equalizer may divide and output the output signal of the subtractor as a basic component of the channel response output from the first delay unit. The equalizer may be implemented as a one-tap equalizer.

Compared to the prior art which requires a large amount of computation and a high computational complexity to remove the ICI component included in the received signal, the ICI reduction equalizer according to the present invention has a relatively low computational complexity and a low computational complexity. Require.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2A and 2B illustrate an ICI reduction equalizer according to a preferred embodiment of the present invention.

2A and 2B show an RF receiver 222, a CP remover 223, and an FFT block 224 with an ICI reduction equalizer 230. The ICI reduction equalizer 230 includes a channel estimator 240, a channel calculator 250, an ICI estimator 260, a subtractor 270, and an equalizer 280. The ICI estimator 260 illustrated in FIGS. 2A and 2B includes a multiplier 261, a filter bank 265, and an adder 269.

)을 추정한다. 도 2a에는 채널 추정부(240)가 주파수 영역의 수신 신호 R_i(k)로부터 채널 응답을 추정하는 실시예가 도시되어 있고, 도 2b에는 채널 추정부(240)가 시간 영역의 수신 신호 r_CP(n)로부터 채널 응답을 추정하는 실시예가 도시되어 있다. r_CP(n)는 CP(Cyclic Prefix)가 제거되기 전의 수신 신호를 나타내고, r_i(n)은 CP가 제거된 수신 신호를 나타낸다. 시간 영역의 수신 신호 r_i(n)는 FFT(Fast Fourier Transform) 블럭(224)에 의하여 주파수 영역의 수신 신호 R_i(k)로 변환된다.The channel estimator 240 performs a channel response from the received signal.

Estimate). 2A illustrates an embodiment in which the channel estimator 240 estimates a channel response from the received signal R _i (k) in the frequency domain. In FIG. 2B, the channel estimator 240 uses the received signal r _CP (in the time domain). An embodiment for estimating the channel response from n) is shown. r _CP (n) represents the received signal before the CP (Cyclic Prefix) is removed, and r _i (n) represents the received signal from which the CP has been removed. The received signal r _i (n) in the time domain is transformed into the received signal R _i (k) in the frequency domain by the Fast Fourier Transform (FFT) block 224.

)으로부터 채널 응답의 기본 성분(basic component.

)과 채널 응답의 변동 성분(fluctuating component.

)을 계산한다. 여기서, 채널 응답의 기본 성분(

)은 1 심볼 구간 동안 변동하지 않는 채널 응답 성분이고, 채널 응답의 변동 성분(

)은 1 심볼 구간 동안 변동하는 채널 응답 성분이다. 채널 응답의 기본 성분(

)과 채널 응답의 변동 성분(

)에 대한 자세한 내용은 도 3을 참조하여 설명한다.The channel calculator 250 estimates the estimated channel response (

From the basic component of the channel response.

) And the fluctuating component of the channel response.

Calculate Where the fundamental component of the channel response (

) Is a channel response component that does not fluctuate during one symbol interval, and the variation component of the channel response (

) Is a channel response component that fluctuates during one symbol period. The fundamental components of the channel response (

) And the variation component of the channel response (

) Will be described in detail with reference to FIG. 3.

)과 주파수 영역의 수신 신호 R_i(k)를 곱하고, 그 곱셈 결과를 필터 계수들 F(q)에 따라 필터링하며, 그 필터링 결과에 기초하여 수신 신호에 포함된 ICI(Inter-Carrier Interference) 성분을 추정한다. ICI 추정부(260)의 곱셈부(261)는 채널 응답의 변동 성분(

)과 주파수 영역의 수신 신호 R_i(k)를 곱하여 그 곱셈 결과를 출력한다. ICI 추정부(260)의 필터 뱅크(265)는 곱셈부(261)로부터 출력되는 곱셈 결과를 필터 계수들 F(q)에 따라 필터링하여 그 필터링 결과를 출력한다. ICI 추정부(260)의 합산부(269)는 필터 뱅크(265)로부터 출력되는 필터링 결과를 합산하여 ICI 성분

를 출력한다. ICI 추정부(260)의 구체적인 내용은 도 3을 참조하여 설명한다.The ICI estimator 260 determines the variation component of the channel response (

Multiply the received signal R _i (k) by the frequency domain, and filter the multiplication result according to the filter coefficients F (q), and based on the filtering result, an inter-carrier interference component included in the received signal Estimate The multiplier 261 of the ICI estimator 260 uses the variation component of the channel response (

) Is multiplied by the received signal R _i (k) in the frequency domain to output the multiplication result. The filter bank 265 of the ICI estimator 260 filters the multiplication result output from the multiplier 261 according to the filter coefficients F (q) and outputs the filtering result. The adder 269 of the ICI estimator 260 sums the filtering result output from the filter bank 265 to add the ICI component.

Outputs Details of the ICI estimator 260 will be described with reference to FIG. 3.

를 감산한다. 등화부(280)는 감산부(270)의 출력 신호를 채널 응답의 기본 성분(

)에 기초하여 등화시킨다. 등화부(280)는 등화 결과를 추정된 송신 신호

로 출력한다. The subtraction unit 270 is an ICI component in the received signal R _i (k) in the frequency domain.

Subtract The equalizer 280 transmits the output signal of the subtractor 270 to the basic component of the channel response (

Equalization). The equalizer 280 transmits the estimated result of the equalization

Will output

를 추정하고, 그 추정 결과에 기초하여 수신 신호에 포함된 ICI 성분

를 제거 또는 저감시키며, ICI 성분

가 제거 또는 저감된 신호를 등화시켜 추정된 송신 신호

를 출력한다. The ICI reduction equalizer 230 shown in FIGS. 2A and 2B may be applied to a receiver of an orthogonal frequency division multiplexing (OFDM) system. That is, when the transmission signal S _i (n) output from the transmitter of the OFDM system is received by the receiver of the OFDM system through a wired or wireless channel, the ICI reduction equalizer 230 included in the receiver of the OFDM system is included in the received signal. ICI Ingredients

Is estimated, and the ICI component included in the received signal based on the estimation result

Removes or reduces

Is estimated by equalizing the removed or reduced signal

Outputs

3 is a diagram specifically illustrating the ICI reduction equalizer 230 shown in FIGS. 2A and 2B.

The ICI reduction equalizer illustrated in FIG. 3 includes a channel estimator 340, a channel calculator 350, first multipliers 361_1 to M1 multipliers 361_M1, and first filters 365_1 to M1 filters 365_M1. And an adder 369, a subtractor 370, and an equalizer 380. The first multipliers 361_1 to M1 multipliers 361_M1 in FIG. 3 correspond to the multipliers 261 in FIGS. 2A and 2B, and the first filters 365_1 to M1 filters 365_M1 in FIG. 3. ) Corresponds to the filter bank 265 in FIGS. 2A and 2B, and the adder 369 in FIG. 3 corresponds to the adder 269 in FIGS. 2A and 2B. In the following, the ICI reduction equalizer illustrated in FIG. 3 will be described with reference to a number of equations.

The transmission signal S _i (n) output from the transmitter of the OFDM system may be represented by Equation 1.

는 예컨대 i 번째 QAM-맵핑 심볼(Quadrature Amplitude Modulation-mapping symbol)을 나타낸다.In Equation 1, Ng is a size of a cyclic prefix (CP), and N is an FFT size (that is, a total number of sub-carriers). And,

Denotes, for example, the i-th QAM-mapping symbol (Quadrature Amplitude Modulation-mapping symbol).

If the discrete response of a multi-path channel, linear time variant, is h _ltv (n, l), then the path is delayed by l times the sampling period. The channel response h _i (n, l) may be defined as in Equation 2.

In Equation 2, P represents the maximum delay among delays caused by the multipath.

On the other hand, if the transmission signal S _i (n) is expressed as Equation 1 and the channel response h _i (n, l) is defined as Equation 2, the received signal r _i (n with the CP removed from the receiver of the OFDM system) ) Is expressed as in Equation 3.

은 AWGN(Additional White Gaussian Noise)을 나타낸다.In equation (3)

Denotes AWGN (Additional White Gaussian Noise).

The received signal r _i (n) in the time domain is converted into the received signal R _i (k) in the frequency domain as shown in Equation 4 by FFT (Fast Fourier Transform).

In Equation 4,

는

를 FFT한 신호이다.ego,

Is

Is a signal obtained by FFT.

Referring to Equation 4, the received signal in the frequency domain may be expressed in a matrix form as shown in Equation 5 below.

In Equation 5, the matrix G _i is a channel gain matrix (or equalization matrix). When orthogonality between subcarriers is maintained, all elements other than the main diagonal elements of the matrix G _i are zero. However, if orthogonality between subcarriers is impaired and an ICI (Inter-Carrier Interference) component is included in the received signal, elements other than the main diagonal elements of the matrix G _i become non-zero. If elements other than the main diagonal elements of the matrix G _i are non-zero, it becomes more complicated for the OFDM receiver to estimate the transmitted signal.

을 배제한다면, 다음의 수학식 6과 같이 심볼을 추정할 수 있다.AWGN component in Equation 5

If it is excluded, the symbol can be estimated as shown in Equation 6 below.

In order to simplify the equalization matrix G _i , the channel is approximated to an M1-order model in Equation 8 below.

을 수학식 7과 같이 정의한다.First, the channel vector for the i th symbol period

Is defined as in Equation 7.

의 각 요소들은 채널 응답 h_ltv(n,l)로부터 추정된 M 개의 샘플값이다.

의 각 요소들은 도 3에서의

에 대응된다. 수학식 7을 참조하면 M1 차 모델로 근사화된 채널은 다음의 수학식 8과 같이 표현될 수 있다. 즉, 채널 응답 h_i(n,l)은 1 심볼 구간 동안 변동하지 않는(즉, time-invariant) 기본 성분(basic component)과 1 심볼 구간 동안 변동하는(즉, time-variant) 변동 성분(fluctuating component)으로 구분될 수 있다.Channel Vector in Equation 7

Each element of M is M sample values estimated from the channel response h _ltv (n, l).

Each element of

Corresponds to. Referring to Equation 7, a channel approximated by the M1 order model may be expressed as Equation 8 below. That is, the channel response h _i (n, l) is a basic component that does not fluctuate during one symbol interval (ie, time-invariant) and fluctuating that fluctuates during one symbol interval (ie, time-variant). component).

In Equation 8,

과 채널 계수

에 의해서 결정된다. 본 발명의 어느 한 실시예에서, i 번째 심볼 구간에 대하여 채널 계수

를 상수(constant)로 설정할 수 있다. 또한, 본 발명의 어느 한 실시예에서,

은 수학식 9를 만족하도록 설정될 수 있다.to be. As shown in Equation 8, the M1 order model of the channel

And channel coefficients

Determined by In one embodiment of the present invention, channel coefficients for the i th symbol interval

Can be set to a constant. Further, in one embodiment of the present invention,

May be set to satisfy the equation (9).

은 채널 응답의 변동 성분(fluctuating component)에 관계된다. 수학식 8에 표현된 채널 응답의 변동 성분(fluctuating component)에서 인자(argument) n과 인자(argument) l은 서로 분리되어 있다. 한편, 채널 응답의 기본 성분(basic component)은 채널 응답 h_i(n,l)의 대부분 에너지(decisive energy)를 품는다. As shown in Equation 8,

Is related to the fluctuating component of the channel response. Argument n and argument l are separated from each other in the fluctuating component of the channel response represented by Equation (8). On the other hand, the basic component of the channel response contains most of the energy of the channel response h _i (n, l).

에 관한 식에 대입하면 수학식 10을 얻는다.Equation 8 in Equation 4

Substituting the equation for, yields equation (10).

은 Kronecker delta(즉, unit delta function)를 나타내고, In Equation 10,

Represents a Kronecker delta (i.e. unit delta function),

이며,

Is,

이다. 수학식 10은 다음의 수학식 11과 같은 행렬 형태로 쓸 수 있다.

to be. Equation 10 may be written in a matrix form as in Equation 11 below.

In Equation 11,

는 N×N의 대각 행렬(diagonal matrix)이다.to be. In other words,

Is a diagonal matrix of N × N.

는 다음과 같다. In addition, the filter coefficient matrix

Is as follows.

는 심볼 인덱스 i에 대하여 독립적(independent)이며 Toplitzian 특성(Toplitzian property)을 가진다. As the determinant above shows, the filter coefficient matrix

Is independent of symbol index i and has a Toplitzian property.

Substituting Equation 11 into Equation 6 yields Equation 12 below.

은 N×N의 단위 행렬(identity matrix)을 나타내고,In Equation 12,

Denotes an identity matrix of N × N,

이다.

는 정규화된 변동 성분(fluctuating component)에 대응된다.

to be.

Corresponds to a normalized fluctuating component.

In order to simplify the equation (12), the high-order component is discarded to obtain the following equation (13).

도 매우 느리게 변한다. 따라서,

을 FFT한

은 저역 통과 특성을 보이게 된다. 본 발명에서는 이와 같은 특성을 감안하여, N 개의 서브 캐리어들 중에서 2 Qp개의 서브 캐리어들에 의한 성분들만을 고려한다. 즉, 캐리어 인덱스가 -Qp ~ +Qp인 서브 캐리어들에 의한 성분들만을 고려한다. 이러한 점을 반영하면, 수학식 13으로부터 다음의 수학식 14가 도출된다.Because the channel response changes very slowly during the symbol period

Also changes very slowly. therefore,

FFT

Has low pass characteristics. In view of the above characteristics, the present invention considers only components of 2 Qp subcarriers among N subcarriers. That is, only components by subcarriers whose carrier indexes are -Qp to + Qp are considered. Reflecting this point, the following equation (14) is derived from equation (13).

Hereinafter, a preferred embodiment of the present invention will be described with reference to Equation 14 and FIG.

또는

)으로부터 채널 응답의 기본 성분(

)과 채널 응답의 변동 성분(

)을 계산한다. 수학식 8에서

에 관한 식의 양변을 FFT하면 수학식 15를 얻는다.The channel calculator 350 estimates the channel response inputted from the channel estimator 340.

or

From the fundamental components of the channel response (

) And the variation component of the channel response (

Calculate In equation (8)

FFT both sides of the equation with respect to equation (15).

로부터

(0≤p≤M1)를 계산한다.

(p=0)는 채널 응답의 기본 성분(

)으로서 등화부(380)로 출력된다. 또한, 채널 계산부(350)는 수학식 12에서

와

의 관계식을 이용하여

로부터

(1≤p≤M1)를 계산한다. 도 3에 도시된 바와 같이, 채널을 M1 차 모델(M1-order model)로 근사화하는 경우에, 채널 계산부(350)는 채널 응답의 제 1 차 변동 성분(

) 내지 채널 응답의 제 M1 차 변동 성분(

)을 출력한다.The channel calculator 350 uses Equation 15

from

Calculate (0 ≦ p ≦ M1).

(p = 0) is the fundamental component of the channel response (

) Is output to the equalization unit 380. In addition, the channel calculator 350 may be

Wow

Using the relation of

from

(1 ≦ p ≦ M1) is calculated. As shown in FIG. 3, when approximating a channel to an M1-order model, the channel calculator 350 may determine the first order variation component of the channel response.

) To the Mth order variation component of the channel response (

)

)과 주파수 영역의 수신 신호 R_i(k)를 곱하여 제 1 곱셈 결과

를 출력한다. 비슷하게, 제 M1 곱셈기(361_M1)는 채널 응답의 제 M1 차 변동 성분(

)과 주파수 영역의 수신 신호 R_i(k)를 곱하여 제 M1 곱셈 결과

를 출력한다.In FIG. 3, the first multiplier 361_1 may use the first order variation component of the channel response (

) And the first multiplication result by multiplying the received signal R _i (k) in the frequency domain

Outputs Similarly, the M1 multiplier 361_M1 is equal to the M < th >

M1 multiplication result by multiplying the received signal R _i (k) by the frequency domain

Outputs

를 제 1 필터 계수들

에 따라 필터링하여 제 1 필터링 결과(I1)를 출력한다. 비슷하게, 제 M1 필터(365_M1)는 제 M1 곱셈기(361_M1)로부터 출력되는 제 M1 곱셈 결과

를 제 M1 필터 계수들

에 따라 필터링하여 제 M1 필터링 결과를 출력한다. 본 발명의 어느 한 실시예에서, 제 1 필터(365_1) 내지 제 M1 필터(365_M1)는 FIR(Finite Impulse Response) 필터로 구현될 수 있다. 예컨대, 필터 뱅크(265)에 구비되는 각 필터들(365_1, 365_2, ..., 365_M1)은 수학식 14에서In FIG. 3, the first filter 365_1 receives the first multiplication result output from the first multiplier 361_1.

First filter coefficients

The first filtering result I1 is output by filtering according to the method. Similarly, the M1 filter 365_M1 outputs the M1 multiplication result output from the M1 multiplier 361_M1.

M1 filter coefficients

Filter according to output the M1 th filtering result. In one embodiment of the present invention, the first filter 365_1 to M1 filter 365_M1 may be implemented as a finite impulse response (FIR) filter. For example, each of the filters 365_1, 365_2,..., 365_M1 included in the filter bank 265 may be represented by Equation 14 below.

를 계산한다. 이하에서는 도 4를 참조하여 FIR 필터를 설명한다.

Calculate Hereinafter, the FIR filter will be described with reference to FIG. 4.

4 is a diagram illustrating the first filter 365_1 in FIG. 3.

와 제 1 필터 계수들[F₁(-Qp), F₁(-Qp+1), F₁(-Qp+2), F₁(-Qp+3), ..., F₁(Qp-1), F₁(Qp)]을 입력받아 FIR 필터링을 수행한다. 합산기(464)는 제 1 필터링 결과(I1)를 도 3에서의 합산부(369)로 출력한다.The first filter 365_1 implemented as the FIR filter in FIG. 4 includes a plurality of delayers 462_1 to 462_2Qp-1, a plurality of multipliers 463_1 to 463_2Qp, and a summer 464. The first filter 365_1 is the first multiplication result

And first filter coefficients F ₁ (-Qp), F ₁ (-Qp + 1), F ₁ (-Qp + 2), F ₁ (-Qp + 3), ..., F ₁ (Qp- 1), F ₁ (Qp)] is input to perform FIR filtering. The summer 464 outputs the first filtering result I1 to the summer 369 in FIG. 3.

내지 제 M1 필터 계수들

는 선형 시불변(linearly time-invariant)이다. 또한, 본 발명의 어느 한 실시예에서, 필터 뱅크(265)에 구비되는 각 필터들의 필터 계수들은 서로 같도록 설정될 수도 있고 서 로 다르도록 설정될 수도 있다. 한편, 제 1 필터(365_1) 내지 제 M1 필터(365_M1) 각각의 필터 탭(tap) 수는 시스템 요구 사양(system requirement)에 따라서 결정된다. 예컨대, 각각의 필터 탭 수는 서로 같을 수도 있고, 각각의 필터 탭 수는 서로 다를 수도 있다. 더 구체적으로, OFDM 수신기의 이동도(mobility)가 높을수록 더 많은 탭 수를 가지는 FIR 필터가 필요하다. Although FIG. 4 illustrates the first filter 365_1 in FIG. 3, the second filter 365_2 to the M1 filter 365_M1 in FIG. 3 may also have the configuration as shown in FIG. 4. In an embodiment of the present invention, the first filter 365_1 to the M1 filter 365_M1 are linear time-invariant filters. That is, the first filter coefficients in FIG.

To M1st filter coefficients

Is linearly time-invariant. In addition, in one embodiment of the present invention, the filter coefficients of the filters provided in the filter bank 265 may be set to be the same or different from each other. On the other hand, the number of filter taps of each of the first filter 365_1 to the M1 filter 365_M1 is determined according to a system requirement. For example, the number of filter taps may be the same as each other, and the number of filter taps may be different from each other. More specifically, the higher the mobility of the OFDM receiver requires a FIR filter having a higher number of taps.

를 출력한다. 예컨대, 합산부(369)는 수학식 14에서In FIG. 3, the adder 369 may filter the first filtering result I1 output from the first filter 365_1 of the filter bank 265 or the M1 filtering output from the M1 filter 365_M1 of the filter bank 265. ICI component by adding up the results

Outputs For example, the adder 369 is expressed by Equation 14

를 계산한다.

Calculate

를 감산한다. 예컨대, 감산부(370)는 수학식 14에서In FIG. 3, the subtractor 370 is an ICI component in the received signal R _i (k) in the frequency domain.

Subtract For example, the subtraction unit 370 is expressed by Equation 14

를 계산한다.

Calculate

)에 기초하여 등화시킨다. 본 발명의 어느 한 실시예에서, 등화부(380)는 1 탭 등화기(one-tap equalizer)로 구현될 수 있다. 1 탭 등화기로 구현되는 등화부(380)는, 감산부(370)의 출력 신호를 채널 응답의 기본 성분(

)으로 제산(dividing)하고, 그 제산 결과를 추정된 송신 신호

로서 출력한다. 예컨대, 1 탭 등화기로 구현되는 등화부(380)는 수학식 14에서In FIG. 3, the equalizer 380 outputs the output signal of the subtractor 370 to the basic component of the channel response.

Equalization). In one embodiment of the invention, the equalizer 380 may be implemented as a one-tap equalizer. The equalizer 380, which is implemented as a one-tap equalizer, outputs the output signal of the subtractor 370 to the basic component of the channel response (

And dividing the result of the division into the estimated transmission signal.

Output as. For example, the equalizer 380 implemented as a one tap equalizer is represented by Equation 14

를 계산한다.

Calculate

번 정도의 복소 곱셈(complex multiplication)이 수행된다. 수신 신호에 포함된 ICI 성분을 제거하기 위해서 많은 연산량과 높은 계산 복잡도(computational complexity)를 요구하는 종래 기술에 비하여, 본 발명에 따른 ICI 저감 등화기는 상대적으로 적은 연산량과 낮은 계산 복잡도(computational complexity)를 요구한다고 볼 수 있다.In the ICI reduction equalizer according to the present invention for estimating and removing the ICI component included in the received signal as described above, approximately

One or more complex multiplications are performed. Compared to the prior art which requires a large amount of computation and a high computational complexity to remove the ICI component included in the received signal, the ICI reduction equalizer according to the present invention has a relatively low computational complexity and a low computational complexity. It may be required.

Meanwhile, in FIG. 3, the first multiplier 361_1 and the first filter 365_1 form a first filtering path, and the second multiplier 361_2 and the second filter 365_2 form a second filtering path. Similarly, the M1 multiplier 361_M1 and the M1 filter 365_M1 form a M1 filtering path. As can be seen from FIG. 3, in the case of approximating a channel to an M1-order model, an ICI reduction equalizer according to a preferred embodiment of the present invention includes M1 filtering paths. In other words, when the channel is approximated with the primary model, one filtering path is provided. When the channel is approximated with the secondary model, two filtering paths are provided. When the channel is approximated with the M1 order model, A filtering path is provided. An example of approximating the channel to the primary model is illustrated in FIG. 5, which will be described in detail with reference to FIG. 5.

Based on the above description, the ICI estimation method according to the preferred embodiment of the present invention will be described.

또는

)으로부터 채널 응답의 제 1 차 변동 성분(

) 내지 채널 응답의 제 M1 차 변동 성분(

)을 계산한다. 그리고, 채널 응답의 제 1 차 변동 성분(

) 내지 채널 응답의 제 M1 차 변동 성분(

) 각각과 주파수 영역의 수신 신호 R_i(k)를 곱하여 제 1 곱셈 결과

내지 제 M1 곱셈 결과

를 출력한다. 그리고, 제 1 곱셈 결과

내지 제 M1 곱셈 결과

각각을 제 1 필터 계수들

내지 제 M1 필터 계수들

각각에 따라 FIR(Finite Impulse Response) 필터링하여 제 1 필터링 결과(I1) 내지 제 M1 필터링 결과를 출력한다. 그리고, 제 1 필터링 결과(I1) 내지 제 M1 필터링 결과를 합산하여 수신 신호에 포 함된 ICI(Inter-Carrier Interference) 성분

를 추정한다. Estimated channel response from received signal (

or

From the first order variation component of the channel response

) To the Mth order variation component of the channel response (

Calculate And the first order variation component of the channel response (

) To the Mth order variation component of the channel response (

) And the first multiplication result by multiplying each received signal R _i (k) in the frequency domain.

To M1th multiplication result

Outputs And the first multiplication result

To M1th multiplication result

Each of the first filter coefficients

To M1st filter coefficients

According to each of the FIR (Finite Impulse Response) filtering to output the first filtering result (I1) to M1 filtering results. The first filtering result (I1) to the first filtering result M1 and the ICI (Inter-Carrier Interference) component included in the received signal

Estimate

또는

)은 1 심볼 구간 동안 변동하지 않는 기본 성분(basic component)과 1 심볼 구간 동안 변동하는 제 1 차 변동 성분(first-order fluctuating component) 내지 제 M1 차 변동 성분(M1th-order fluctuating component)으로 구분될 수 있다. As described above, when approximating the channel to the M1-order model, the channel response estimated from the received signal (

or

) May be divided into a basic component that does not fluctuate during one symbol period and a first-order fluctuating component that fluctuates during one symbol period to a M1th-order fluctuating component. Can be.

5 is a diagram illustrating an ICI reduction equalizer in the case of approximating a channel to a primary model according to a preferred embodiment of the present invention.

The ICI reduction equalizer shown in FIG. 5 includes a channel estimator 540, a channel calculator 550, a first multiplier 561, a first filter 565, a subtractor 570, and an equalizer 580. do. In FIG. 5, the channel calculator 550 includes a first delayer 551, a second delayer 553, a subtractor 555, and a divider 557. Since only one filtering path is provided when approximating the channel to the primary model, the ICI reduction equalizer shown in FIG. 5 does not require components such as the adder 369 in FIG.

또는

)을 추정한다. 이하에서는 도 6을 참조하여, 채널을 1 차 선형 모델(1-order linear model)로 근사화하는 경우에, 채널 추정부(540)에 의한 채널 응답의 추정을 설명한다.The channel estimator 540 receives a channel response from the received signal.

or

Estimate). Hereinafter, referring to FIG. 6, the estimation of the channel response by the channel estimator 540 when the channel is approximated by a 1-order linear model will be described.

6 is a diagram illustrating a channel response in an i-1 th symbol period, an i th symbol period, and an i + 1 th symbol period.

As illustrated in FIG. 6, the channel estimator 540 may take a sample value of channel response at center of the symbol interval. That is, the channel response sample value may be taken at the (N / 2) -1 point in the symbol period from 0 to N-1.

에 기초하여 이전 심볼(previous symbol) 구간에서의 채널 응답 추정값

을 출력하고, i 번째 심볼 구간 중앙에서의 채널 응답 샘플값

에 기초하여 현재 심볼(present symbol) 구간에서의 채널 응답 추정값

를 출력하며, i+1 번째 심볼 구간 중앙에서의 채널 응답 샘플값

에 기초하여 다음 심볼(next symbol) 구간에서의 채널 응답 추정값

을 출력한다. 당업자라면,

은

에 대응되고,

는

에 대응되며,

는

에 대응된다는 점을 이해할 것이다.Assuming that the i th symbol is a present symbol, the channel response sample value in the i-1 th symbol interval to approximate the channel to the 1-order linear model for the i th symbol interval, Consider the case of using the channel response sample value in the i th symbol period and the channel response sample value in the i + 1 th symbol period. In this case, the channel estimator 540 determines the channel response sample value at the center of the i-1 th symbol interval.

Estimation of channel response in previous symbol interval based on

Outputs the channel response sample value at the center of the i th symbol interval.

Channel response estimate in the current symbol interval based on

Channel response sample value at the center of the i + 1 th symbol interval

Channel response estimate in the next symbol interval based on

Outputs Those skilled in the art

silver

Corresponding to

Is

Corresponds to

Is

Will be understood.

, i 번째 심볼 구간 중앙에서의 채널 응답 샘플값

및 i+1 번째 심볼 구간 중앙에서의 채널 응답 샘플값

을 이용하여 i 번째 심볼 구간에 대하여 채널을 1 차 선형 모델(1-order linear model)로 근사화하는 경우에, i 번째 심볼 구간에 대한 채널 벡터를 정의하는 수학식 7은 다음의 수학식 16과 같이 표현될 수 있다. 수학식 7에서의 M은 수학식 16에서 3이다.Channel response sample value in the center of i-1 th symbol interval

channel response sample at the center of i-th symbol interval

And channel response sample value at the center of i + 1 th symbol interval

In the case of approximating a channel with a 1-order linear model with respect to the i-th symbol interval using, Equation 7 defining a channel vector for the i-th symbol interval is expressed by Equation 16 below. Can be expressed. M in (7) is 3 in (16).

In order to approximate the channel to the 1-order linear model, when M1 is set to 1 in Equation 8, the following Equation 17 is obtained.

에 관한 식에 포함된 채널 계수

를 다음의 수학식 18과 같이 설정하면, 아래의 수학식 19가 얻어진다.Meanwhile, in Equation 8

Channel coefficients included in equations for

Is set as in Equation 18, the following Equation 19 is obtained.

에 관한 식을 참조하면,

을 FFT한 신호의 MMSE(Minimum-Mean-Squared-Error) 추정은 수학식 20과 같이 표현될 수 있다.In equation (19)

If you refer to the equation for

Minimum-Mean-Squared-Error (MMSE) estimation of the signal obtained by FFT may be expressed as in Equation 20.

에 관한 식을 이용하면 채널 응 답의 제 1 차 변동 성분

은 다음의 수학식 21과 같이 계산될 수 있다.In equation (19), equation (20) and equation (12)

Using the equation for, the first order variation component of the channel response

May be calculated as in Equation 21 below.

에 관한 식을 참조하면, 채널 응답의 기본 성분은 수학식 22와 같이 표현될 수 있다. 여기서, 채널 응답의 기본 성분은 i-1 번째 심볼 구간, i 번째 심볼 구간 및 i+1 번째 심볼 구간에서의 채널 응답 평균값이라고 볼 수 있다.On the other hand, in equation (19)

Referring to the equation for, the basic component of the channel response can be expressed as Equation 22. Here, the basic component of the channel response may be regarded as an average value of the channel response in the i-1 th symbol period, the i th symbol period, and the i + 1 th symbol period.

Based on Equations 16 to 22, Equation 14 is obtained by rewriting Equation 14 to fit the case of approximating a channel to a 1-order linear model.

을 FFT하면 수학식 23에서의

가 얻어진다. 도 5에서 보듯이,

는 제 1 필터(565)로 입력되는 제 1 필터 계수들이다. i 번째 심볼 구간에 대하여 채널을 1 차 선형 모델(1-order linear model)로 근사화하는 경우에

은 다음의 수학식 24와 같이 설정될 수 있다.In equation (17)

FFT is given by Equation 23

Is obtained. As shown in Figure 5,

Are first filter coefficients input to the first filter 565. In the case of approximating the channel to the 1-order linear model for the i th symbol interval,

May be set as in Equation 24 below.

Hereinafter, the ICI reduction equalizer shown in FIG. 5 will be described with reference to Equation 23. FIG.

, 현재 심볼 구간에서의 채널 응답 추정값

및 다음 심볼 구간에서의 채널 응답 추정값

을 이용하여 채널 응답의 기본 성분(basic component)

과 채널 응답의 제 1 차 변동 성분(first-order fluctuating component)

을 계산한다. 앞서 설명한 바 있듯이, 채널 응답의 기본 성분

은 1 심볼 구간 동안 변동하지 않는 채널 응답 성분이고, 채널 응답의 제 1 차 변동 성분

은 1 심볼 구간 동안 변동하는 채널 응답 성분이다. In FIG. 5, the channel calculator 550 estimates the channel response in the previous symbol interval.

, The channel response estimate in the current symbol interval

And channel response estimates in the next symbol interval

The basic component of the channel response using

First-order fluctuating component of the channel response

Calculate As mentioned earlier, the fundamental components of the channel response

Is a channel response component that does not vary during one symbol interval and is the first order variation component of the channel response.

Is a channel response component that fluctuates during one symbol period.

를 채널 응답의 기본 성분

으로서 출력한다. 또한, 채널 계산부(550)는, 다음 심볼 구간에서의 채널 응답 추정값

에서 이전 심볼 구간에서의 채널 응답 추정값

을 감산하고, 그 감산 결과를 현재 심볼 구간에서의 채널 응답 추정값

로 제산(dividing)하며, 그 제산 결과를 채널 응답의 제 1 차 변동 성분

으로서 출력한다. 즉, 채널 계산부(550)는 수학식 22의 연산과 수학식 21의 연산을 담당한다.As shown in FIG. 5, the channel calculator 550 estimates a channel response in the current symbol interval.

Basic components of channel response

Output as. In addition, the channel calculator 550 estimates the channel response in the next symbol period.

Channel Response Estimation in Previous Symbol Interval in

And subtract the result of the subtraction into the channel response estimate in the current symbol interval.

Dividing by, and divides the result of the first order variation of the channel response.

Output as. That is, the channel calculator 550 is responsible for the calculation of Equation 22 and the calculation of Equation 21.

에서 제 2 지연기(553)로부터 출력되는 이전 심벌 구간에서의 채널 응답 추정값

을 감산한다. 채널 계산부(550)에 구비되는 제산기(557)는 감산기(555)로부터 출력되는 감산 결과를 제 1 지연기(551)로부터 출력되는 현재 심볼 구간에서의 채널 응답 추정값

으로 제산하고, 그 제산 결과를 채널 응답의 제 1 차 변동 성분

으로서 제 1 곱셈기(561)로 출력한다. For the calculation of Equation 22 and the calculation of Equation 21, the first delay unit 551 included in the channel calculator 550 delays the channel response estimate output from the channel estimator 540 for one symbol period. And print it out. The second delay unit 553 included in the channel calculator 550 delays the output signal of the first delay unit 551 for one symbol period and outputs the delayed signal. The subtractor 555 included in the channel calculator 550 estimates a channel response in the next symbol period output from the channel estimator 540.

Estimates the channel response in the previous symbol period output from the second delayer 553 in

Subtract The divider 557 included in the channel calculator 550 estimates the channel response in the current symbol interval output from the first delayer 551 by subtracting the output result from the subtractor 555.

Divide the result of the division by the first order variation component of the channel response.

As a first multiplier 561.

과 주파수 영역의 수신 신호 R_i(k)를 곱하여 제 1 곱셈 결과

를 출력한다. In Figure 5 the first multiplier 561 is the first order variation component of the channel response.

And the first multiplication result by multiplying the received signal R _i (k) in the frequency domain by

Outputs

를 제 1 필터 계수들

에 따라 FIR(Finite Impulse Response) 필터링하여 수신 신호에 포함된 ICI(Inter-Carrier Interference) 성분

를 출력한다. 예컨대, 제 1 필터(565) 는 수학식 23에서 In FIG. 5, the first filter 565 has a first multiplication result.

First filter coefficients

Component based on IIC (Finite Impulse Response) filtering, which is included in the received signal

Outputs For example, the first filter 565 may be

를 계산한다.

Calculate

를 감산한다. 예컨대, 감산부(570)는 수학식 23에서 In FIG. 5, the subtractor 570 may include an ICI component in the received signal R _i (k) in the frequency domain.

Subtract For example, the subtraction unit 570 is expressed by Equation 23

를 계산한다.

Calculate

)에 기초하여 등화시킨다. 도 5에서의 등화부(580)는 1 탭 등화기(one-tap equalizer)로 구현될 수 있다. 1 탭 등화기로 구현되는 등화부(580)는, 감산부(570)의 출력 신호를 제 1 지연기(551)로부터 출력되는 채널 응답의 기본 성분(

)으로 제산(dividing)하고, 그 제산 결과를 추정된 송신 신호

로서 출력한다. 예컨대, 1 탭 등화기로 구현되는 등화부(580)는 수학식 23에서In FIG. 5, the equalizer 580 transmits the output signal of the subtractor 570 to the basic component of the channel response.

Equalization). The equalizer 580 in FIG. 5 may be implemented as a one-tap equalizer. The equalizer 580, which is implemented as a one-tap equalizer, may use the basic component of the channel response output from the first delay unit 551 to output the output signal of the subtractor 570.

And dividing the result of the division into the estimated transmission signal.

Output as. For example, the equalizer 580 implemented as a one tap equalizer is represented by Equation 23.

를 계산한다.

Calculate

FIG. 7 is a simulation graph showing the relationship between the maximum Doppler shift fd and the minimum SNR value SNRmin when the ICI reduction equalizer shown in FIG. 5 is applied to a receiver of a digital video broadcasting-handheld (DVB-H) system. .

FIG. 8 is a simulation graph showing the relationship between the maximum Doppler shift fd and the packet error rate PER when the ICI reduced equalizer shown in FIG. 5 is applied to a receiver of a DVB-H system.

In FIG. 7 and FIG. 8, simulation conditions are as follows. In the DVB-H system, the FFT size was 8k, the Guard Interval Ratio was 1/4, 16 Quadrature Amplitude Modulation (QAM) modulation schemes were used, and the COST207 TU6 channel was tested. As shown in FIG. 7 and FIG. 8, the number of filter taps Q of the first filter 565 in FIG. 5 is set to five.

In FIG. 7, SNRmin means a minimum Signal to Noise Ratio (SNR) value such that a packet error rate is 5% or less. If the SNRmin is set to 21.5 dB, the performance gain is 40 Hz when performing ICI reduction according to the present invention (ICI mitigation with Q = 5) compared to not performing ICI reduction (no ICI mitigation). Increases.

FIG. 8 illustrates a packet error rate PER according to the maximum Doppler shift fd when the SNR value is set to 50 dB. Referring to the case where the packet error rate (PER) is 5% in FIG. 8, the performance gain in the case of performing ICI reduction according to the present invention (ICI mitigation with Q = 5) compared to the case in which no ICI reduction is performed (no ICI mitigation). It can be seen that the performance gain increases by 40 Hz.

In the above described the present invention with reference to the specific embodiment shown in the drawings, but this is merely illustrative. Those skilled in the art will appreciate that various modifications and variations are possible therefrom. Therefore, the protection scope of the present invention should be interpreted by the claims to be described later, and all the technical ideas within the equivalent and equivalent ranges should be construed as being included in the protection scope of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS In order to understand the drawings referred to in the detailed description of the invention, a brief description of each drawing is provided.

1 is a diagram illustrating a transmitter and a receiver in an OFDM system.

2A and 2B illustrate an ICI reduction equalizer according to a preferred embodiment of the present invention.

3 is a diagram specifically illustrating the ICI reduction equalizer 230 shown in FIGS. 2A and 2B.

4 is a diagram illustrating the first filter 365_1 in FIG. 3.

5 is a diagram illustrating an ICI reduction equalizer in the case of approximating a channel to a primary model according to a preferred embodiment of the present invention.

6 is a diagram illustrating a channel response in an i-1 th symbol period, an i th symbol period, and an i + 1 th symbol period.

FIG. 7 is a simulation graph showing the relationship between the maximum Doppler shift fd and the minimum SNR value SNRmin when the ICI reduction equalizer shown in FIG. 5 is applied to a receiver of a digital video broadcasting-handheld (DVB-H) system. .

FIG. 8 is a simulation graph showing the relationship between the maximum Doppler shift fd and the packet error rate PER when the ICI reduced equalizer shown in FIG. 5 is applied to a receiver of a DVB-H system.

<Description of Reference Number in Drawing>

110: OFDM transmitter 111: encoder

112: mapper 113: IFFT block

114: CP insertion unit 115: RF transmitter

116: transmit antenna 120: OFDM receiver

121: receiving antenna 122, 222: RF receiving unit

123, 223: CP removing unit 124, 224: FFT block

125: equalizer 126: demapper

127: decoder

230: ICI Reduced Equalizer

240, 340, 540: channel estimator

250, 350, 550: channel calculator

551: first delay 553: second delay

555: subtractor 557: divider

260: ICI estimator

261: multiplier 561: first multiplier

361_1 to 361_M1: first to M1 multipliers

265: filter bank 565: first filter

365_1 to 365_M1: first filter to M1 filter

462_1 to 462_2Qp-1: multiple delayers

463_1 through 463_2Qp: multiple multipliers

464: summer

269, 369: summing

270, 370, 570: subtractive part

280, 380, 580: equalizer

Claims (25)

A channel estimator estimating a channel response from the received signal; A channel calculator which calculates a basic component of the channel response and a fluctuating component of the channel response from the estimated channel response; Multiplying the variation component of the channel response by the received signal in the frequency domain, filtering the multiplication result according to filter coefficients, and estimating an inter-carrier interference component included in the received signal based on the filtering result. ICI estimator; A subtraction unit for subtracting the ICI component from the received signal in the frequency domain; And An equalizer for equalizing the output signal of the subtractor based on a basic component of the channel response; ICI reduction equalizer having a. The method of claim 1, The channel estimator, And estimating the channel response from the received signal in the frequency domain or the received signal in the time domain. The method of claim 2, And the received signal in the frequency domain is a signal obtained by fast Fourier transform (FFT) of the received signal in the time domain. The method of claim 1, The basic component of the channel response is a channel response component that does not vary during one symbol interval, and the variation component of the channel response is a channel response component that varies during one symbol interval. The method of claim 1, In the case of approximating the channel to the M1-order model, And the channel calculator outputs the first order variation component of the channel response to the first order variation component of the channel response to the ICI estimator. The method of claim 1, The ICI estimator, A multiplier for multiplying a variation component of the channel response by a received signal in the frequency domain and outputting a multiplication result; A filter bank for filtering the multiplication result according to the filter coefficients and outputting the filtering result; And An adder configured to add the filtering results and output the ICI component; ICI reduction equalizer, characterized in that it comprises a. The method of claim 6, In the case of approximating the channel to the M1-order model, the multiplier is A first multiplier outputting a first multiplication result by multiplying the first order variation component of the channel response output from the channel calculator and the received signal in the frequency domain; To A M1 multiplier for outputting a M1 multiplication result by multiplying the M1th variation component of the channel response output from the channel calculator and the received signal in the frequency domain; ICI reduction equalizer, characterized in that it comprises a. The method of claim 6, In case of approximating the channel to M1-order model, the filter bank is A first filter outputting a first filtering result by filtering the first multiplication result output from the first multiplier of the multiplier according to first filter coefficients; To A M1 filter outputting the M1 filtering result by filtering the M1 multiplication result output from the M1 multiplier of the multiplier according to M1 filter coefficients; ICI reduction equalizer, characterized in that it comprises a. The method of claim 8, ICI reduction equalizer, characterized in that the first filter to the M1 filter is a linear time-invariant filter. The method of claim 9, ICI reduction equalizer, characterized in that the first to M1 filter is a Finite Impulse Response (FIR) filter. The method of claim 10, The number of filter taps of each of the first to M1 filters is determined according to a system requirement. The method of claim 6, In the case of approximating the channel to the M1-order model, the summing unit may include: And summing the first filtering result output from the first filter of the filter bank and the first M1 filtering result output from the first M1 filter of the filter bank to output the ICI component. The method of claim 1, The equalizer is ICI reduced equalizer, characterized in that implemented as a one-tap equalizer. The method of claim 1, ICI reduced equalizer, characterized in that applied to the receiver of an Orthogonal Frequency Division Multiplexing (OFDM) system. Calculating a first order variation component of the channel response to a first order variation component of the channel response from the channel response estimated from the received signal; Multiplying each of the first order variation component of the channel response to the first order variation component of the channel response by a received signal in a frequency domain to output a first multiplication result to a M1 multiplication result; Filtering each of the first multiplication result to the M1 multiplication result according to each of the first filter coefficients to the M1 filter coefficients and outputting the first to M1 filtering results; And Estimating an Inter-Carrier Interference (ICI) component included in the received signal by summing the first filtering result and the M1 filtering result; ICI estimation method comprising a. The method of claim 15, In the case of approximating the channel to the M1-order model, A basic component which does not vary the channel response estimated from the received signal during one symbol period and the first-order fluctuating component that varies during one symbol period to the Mth order variation component. ICI estimation method characterized in that the classification by (M1th-order fluctuating component). The method of claim 15, A first impulse response (FIR) filter is performed on each of the first multiplication result and the M1 multiplication result according to each of the first filter coefficients and the first M1 filter coefficients to output the first to M1 filtering results. ICI estimation method, characterized in that. A channel estimator estimating a channel response from the received signal; A channel calculator for calculating a basic component of the channel response and a first-order fluctuating component of the channel response from the estimated channel response; A first multiplier for multiplying a first order variation component of the channel response by a received signal in a frequency domain to output a first multiplication result; A first filter outputting an inter-carrier interference (ICI) component included in the received signal by filtering the first multiplication result according to finite impulse response (FIR) according to first filter coefficients; A subtraction unit for subtracting the ICI component from the received signal in the frequency domain; And An equalizer for equalizing the output signal of the subtractor based on a basic component of the channel response; ICI reduction equalizer having a. The method of claim 18, The basic component of the channel response is a channel response component that does not vary during one symbol interval, and the first order variation component of the channel response is a channel response component that varies during one symbol interval. The method of claim 18, The channel estimator, output a channel response estimate in a previous symbol interval based on a sample value of channel response at center of the (i-1) th symbol interval; , outputs a channel response estimate value in a present symbol period based on a sample value of channel response at center of the (i) th symbol interval; outputting a channel response estimate value in a next symbol interval based on a sample value of channel response at center of the (i + 1) th symbol interval; ICI reduction equalizer, characterized in that. The method of claim 20, wherein the channel calculator, Output a channel response estimate in the current symbol interval as a basic component of the channel response, Subtract the channel response estimate in the previous symbol interval from the channel response estimate in the next symbol interval, divide the subtraction result by the channel response estimate in the current symbol interval, and divide the result of the division into the channel response. ICI reduction equalizer, characterized in that it is output as a first order variation component. The method of claim 21, wherein the channel calculator, A first delayer for delaying and outputting a channel response estimate value output from the channel estimator for one symbol period; A second delayer for delaying and outputting the output signal of the first delayer for one symbol period again; A subtractor for subtracting the channel response estimate value in the previous symbol interval output from the second delay unit from the channel response estimate value in the next symbol interval output from the channel estimator; And Subtracting the subtraction result output from the subtractor by the channel response estimate value in the current symbol interval output from the first delayer, and outputting the division result to the first multiplier as a first order variation component of the channel response. Antacid; ICI reduction equalizer, characterized in that it comprises a. The method of claim 22, The channel calculation unit ICI reduction equalizer, characterized in that to approximate the channel to the first-order linear model (1-order linear model). The method of claim 22, The equalizing unit, And dividing the output signal of the subtracter by the basic component of the channel response output from the first delayer. The method of claim 24, The equalizer is ICI reduced equalizer, characterized in that implemented as a one-tap equalizer.

Images